[gmx-users] pdb2gmx eats a bond

Justin A. Lemkul jalemkul at vt.edu
Wed Sep 22 18:28:48 CEST 2010

Thomas Schlesier wrote:
> Hi all,
> i have built an .rtp entry for a molecule with 310 bonds, for the opls 
> force field. For pdb2gmx i use the following command (version 4.0.7)
> *pdb2gmx -f CAT.pdb -o -p -ignh -ter*
> -ter is there because it is no protein, and use no C- and N-terminus 
> (since there are none).
> Her one question, the option -ter works (i think) but it is not listed with
> pdb2gmx -h
> Are there any reasons.

The -ter option is listed for me in 4.0.7:

/usr/local/gmx_407/bin/pdb2gmx -h
-[no]inter   bool   no      Set the next 8 options to interactive
-[no]ss      bool   no      Interactive SS bridge selection
-[no]ter     bool   no      Interactive termini selection, iso charged
-[no]lys     bool   no      Interactive Lysine selection, iso charged

The option certainly works, if that's the concern.

> The relevant output line of pdb2gmx is then
> *Number of bonds was 310, now 309*
> -> So pdb2gmx ate a bond.
> Has anybody an idea why this happens?

I don't know all of the underlying code, but I frequently see a bond removed 
from most of my protein structures.

> Has anybody an idea which could be the fastest way to find out which 
> bond is missing?

Run a short MD and see if any of your atoms drift away :)

> Are the [ angles ] and [ dihedrals ] section also affected by this 
> (since if i understood it correctly pdb2gmx uses the [ bonds ] section 
> to determine the other two sections)?

Potentially, but until you test to see if any bonds are actually missing, 
there's nothing to diagnose in these sections.


> Greetings
> Thomas


Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080


More information about the gromacs.org_gmx-users mailing list